Models & Pricing

Creor gives you access to the best AI models from every major provider through a single interface. Use the Creor Gateway for managed access, or bring your own API keys.

Plans

Creor offers three subscription tiers through the Creor Gateway, plus a bring-your-own-key option that is always free. All plans include the full Creor IDE with all features — plans only differ in how you access AI models.

PlanPriceRequests / MonthBest For
Free$0200 Gateway requestsTrying Creor, light usage
Starter$20/month1,000 Gateway requestsIndividual developers
Pro$40/monthUnlimited Gateway requestsDaily professional use
BYOK$0Unlimited (your own keys)Users who prefer direct API access

Note

All paid plans are billed monthly. You can upgrade, downgrade, or cancel at any time from the dashboard. Unused requests do not roll over.

Free Tier

The Free tier is a great way to try Creor without any commitment. You get 200 Creor Gateway requests per month, which is enough for light exploration and small tasks.

  • 200 Gateway requests per month across all models.
  • Access to all supported models (subject to per-model rate limits).
  • Full IDE features: editor, terminal, extensions, git integration.
  • All 25+ AI tools: file editing, code search, web search, planning, and more.
  • Community support via GitHub Discussions.

When you reach the monthly limit, you can either upgrade to a paid plan or add your own API keys (BYOK) to continue using AI features without limits.

Starter

The Starter plan is designed for individual developers who use AI coding assistance regularly but not all day.

  • 1,000 Gateway requests per month.
  • Priority model access — reduced queue times during peak hours.
  • Email support with 24-hour response time.
  • Usage dashboard with per-session cost breakdown.

A typical coding session uses 10-30 requests depending on task complexity, so 1,000 requests covers roughly 30-100 sessions per month.

Pro

The Pro plan removes all usage limits. It is built for developers who rely on AI assistance throughout their workday.

  • Unlimited Gateway requests — no monthly caps.
  • Access to the latest models on launch day.
  • Priority routing and lower latency.
  • Cloud agents for running AI tasks in the background.
  • Team features: shared workspaces, usage analytics, admin controls.
  • Priority support with same-day response.

BYOK (Bring Your Own Keys)

If you already have API keys from Anthropic, OpenAI, Google, or any other supported provider, you can use them directly in Creor at no cost. BYOK mode is always free and has no request limits — you pay the provider directly at their standard rates.

Setting up BYOK

  • Open Settings (Cmd + , on macOS, Ctrl + , on Windows/Linux).
  • Navigate to AI > Providers.
  • Select a provider (e.g., Anthropic) and enter your API key.
  • The key is stored in your OS keychain — never sent to Creor servers.
  • Models from that provider now appear in the model picker.

Mixing Gateway and BYOK

You can use both the Creor Gateway and your own keys simultaneously. For example, use the Gateway for Claude models and your own OpenAI key for GPT-4o. The model picker shows which provider handles each model.

Tip

BYOK is also useful as a fallback. If the Gateway is temporarily at capacity, your own keys provide uninterrupted access.

Supported Models

Creor supports models from 19+ providers. The following table lists the most popular models available through the Creor Gateway. BYOK users can access any model their provider offers.

Flagship models

ModelProviderContext WindowBest For
Claude Opus 4Anthropic200K tokensComplex reasoning, large codebases
Claude Sonnet 4Anthropic200K tokensFast edits, balanced quality/speed
GPT-4oOpenAI128K tokensGeneral coding, broad knowledge
GPT-4.1OpenAI1M tokensLong-context tasks, large refactors
Gemini 2.5 ProGoogle1M tokensMassive context, multi-file analysis
Gemini 2.5 FlashGoogle1M tokensFast responses, cost-efficient

Additional models

ModelProviderContext Window
Claude Haiku 3.5Anthropic200K tokens
GPT-4o MiniOpenAI128K tokens
Grok 3xAI131K tokens
Mistral LargeMistral128K tokens
Command R+Cohere128K tokens
DeepSeek V3DeepSeek / Together AI64K tokens
Llama 4 MaverickMeta / Groq / Together AI1M tokens
Qwen 3 235BAlibaba / Together AI128K tokens

Provider list

Creor integrates with the following providers. See the Providers documentation for setup details on each one.

  • Anthropic
  • OpenAI
  • Google AI (Gemini)
  • Google Vertex AI
  • AWS Bedrock
  • Azure OpenAI
  • Mistral
  • Groq
  • Cohere
  • xAI (Grok)
  • Perplexity
  • Together AI
  • DeepInfra
  • Cerebras
  • OpenRouter
  • GitLab
  • Vercel AI
  • Fireworks AI
  • Custom OpenAI-compatible endpoints

Credits

Gateway requests are counted per agent interaction, not per API call. When you send a message in the chat, the agent may make multiple internal API calls (reading files, searching, editing) — this counts as a single request from your quota.

How credits are consumed

ActionCredits Used
Chat message (any model)1 credit
Plan mode analysis1 credit
Background agent task1 credit per message
Inline edit suggestion1 credit
Auto-complete (if enabled)0 credits (uses local model)

You can monitor your usage in real time from the dashboard. The status bar in Creor also shows your remaining credits for the current billing period.

Note

Pro plan users have unlimited credits, so usage tracking is informational only — there are no caps or throttles.

Switching Models

You can change the active model at any time, even mid-conversation.

From the model picker

Click the model name in the chat input area (bottom of the AI panel) to open the model picker. It shows all available models grouped by provider, with indicators for speed, cost, and context window size.

From the keyboard

Use the keyboard shortcut Cmd + Shift + M (macOS) or Ctrl + Shift + M (Windows/Linux) to open the model picker without leaving the keyboard.

Per-task model selection

Different models excel at different tasks. A practical setup:

  • Claude Opus 4 or GPT-4.1 — complex architecture decisions, large refactors, debugging tricky issues.
  • Claude Sonnet 4 or GPT-4o — everyday coding, writing tests, quick edits. Best balance of quality and speed.
  • Gemini 2.5 Pro — tasks that need massive context, like analyzing an entire codebase or cross-referencing many files.
  • Claude Haiku 3.5 or GPT-4o Mini — simple tasks, boilerplate generation, formatting. Fastest response times.

Tip

The model picker remembers your last used model per workspace. If you always use Claude Sonnet for your main project and GPT-4o for a side project, Creor switches automatically when you change workspaces.